Fitted IC₅₀ values are shown for the indicated NR1-1/NR2 combinations (assuming complete inhibition for NR2B,C,D); Hill slopes ranged from 0.6 to 1.5. Mean IC₅₀ values were similar to those determined from the composite curve. All experiments were performed with constructs that contain exon 21 and exon 22. For NR2A the inhibition curves was generated using tricine-buffered Zn²⁺ (see Materials and Methods). The mutation in the M2 region (N616R or N637R) enhances the effect of exon 5 on the NR1-1/NR2B and NR1-1/NR2D IC₅₀ values, possibly by removing some Zn²⁺ channel block at high concentrations that probably increased overall Zn²⁺ inhibition. Voltage-dependent and -independent IC₅₀ values for NR2A are different enough such that no voltage-dependent block should contaminate inhibition by nanomolar concentrations of Zn²⁺ (Paoletti et al., 1997). NR2C has very little voltage-dependent inhibition (Chen et al., 1997). Thus the effects of N616R mutation were not determined (ND) for these two subunit combinations. All dose–response curves were determined with five or more Zn²⁺ concentrations; n is the number of oocytes.